) A SIPSCLATICL EVALUATION GP A N87- 19393 CE bITH Ab AUTCEETIC TESMINAL EM (BASA) 21 p CSCL 17ti Unclas H1/06 43501 NASA Tech ni ca I Paper 2669 1987 A Simulation Evaluation of a Pilot Interface With an Automatic Terminal Approach System David A. Hinton Langley Research Center Hampt o n, Virginia National Aeronautics and Space Administration Scientific and Technical Information Branch Summary with a high potential for mistakes and has limited time to detect and correct any errors. A successful A piloted-simulation study was performed to arrival depends on the correct interpretation of ap- evaluate the pilot interface with an automatic termi- proach chart details, the correct setting of numerous nal approach system (ATAS). The ATAS was con- cockpit controls, and precise aircraft guidance near ceived as a concept for improving the pilot inter- the ground. face with high levels of automation. It consists of Automation in the form of an autopilot has been instrument approach data storage, automatic radio used to reduce pilot work load and improve pilot tuning, autopilot, autothrottle, and annunciation of performance in the terminal area. Research studies These components allow the ATAS flight status. (ref. 2) and airplane accident and incident reports to automatically execute instrument approaches, in- suggest, however, that the probability of pilot error cluding procedure turns, altitude changes, missed actually increases with an increase in automation, approaches, and holding patterns, without requir- partially because of design limitations of the pilot- ing the pilot to set up navigation radios or change machine interface. Conventional autopilot interfaces autopilot modes. provide the pilot with many opportunities to make The results show that fewer pilot blunders were errors because of the requirements to change radio made during approaches when using the ATAS than frequencies and autopilot modes as the approach when using a baseline, heading-select autopilot. With progresses. the baseline autopilot, blunders tended to involve loss The automatic terminal approach system (ATAS) of navigational situational awareness or instrument was developed to study ways of significantly reduc- misreading, while with the ATAS, blunders tended to ing the likelihood of pilot error and the work load involve a lack of awareness of the ATAS mode state. during terminal area operations by improving the The ATAS display provided adequate approach- pilot-machine interface. Elements of the ATAS are status information to maintain navigational situa- instrument approach data storage, automatic radio tional awareness, and a side-task measure did not tuning, autopilot, and autothrottle. These elements show any significant difference in pilot work load are used to automatically fly instrument approaches. between the two levels of automation. The ATAS was designed to fly according to the same rules and procedures that the pilot operates by in the Introduction present ATC system. An automatic missed approach General aviation instrument flight rules (IFR) ac- is executed if the pilot does not assume control at tivities currently involve approximately 18 million the missed approach point. Finally, the ATAS was airport operations per year and are forecast by the designed to accommodate the “real world” require- Federal Aviation Administration (FAA) to increase ment of frequent ATC radar vectoring to the final to about 30 million operations per year by 1993. approach course. Most of these operations are conducted by single- A simulated ATAS was developed and a research pilot crews. The air traffic control (ATC) system ex- prototype was built in order to study a concept for pects all aircraft to perform at the same high level. improving the pilot interface with high levels of cock- The high IFR accident rate during the approach pit automation and to evaluate pilot acceptance of and landing phase of flight, as documented in ref- such a system. Details of the simulated ATAS design erence 1, indicates that the required level of perfor- and of its implementation into the Langley General mance has not yet been achieved. It is thought that Aviation Simulator are presented in reference 3. A this level cannot be reached without improvements in piloted-simulation evaluation of the ATAS was then aircraft handling qualities, displays, automatic flight performed. This report presents the results of an control systems, weather information dissemination, evaluation that involved seven instrument-rated pi- training, and ATC procedures. lots, each flying four instrument approaches with The approach and landing is normally the phase the ATAS and four approaches with the baseline of IFR flight with the highest work load. The pi- autopilot. The ATAS was well accepted by the pilots lot must navigate with greater precision than during and the ATAS runs resulted in lower flight technical the departure or en route phases. Air traffic con- error and fewer pilot blunders than with the baseline trol communications and frequency changes are more autopilot. numerous. The pilot must make frequent changes in aircraft altitude, heading, speed, and configura- Abbreviations and Symbols tion while approaching the ground. Checklists must Abbreviations be accomplished and navigation charts frequently re- ferred to. The pilot is in a highly dynamic situation ADF automatic direction finder AKQ identifier for Wakefield Municipal to a general-purpose digital computer. The cockpit is Airport fully enclosed by the cabin section of a light-airplane fuselage. The simulator instrumentation and avion- ATAS automatic terminal approach system ics are typical of an IFR-equipped high-performance ATC air traffic control single-engine or light twin-engine airplane. They in- AUTO ATAS operating mode that uses stored clude an HSI, dual VOR receivers, an ADF, and a approach data to automatically guide two-axis autopilot. An array of speakers provide re- the aircraft alistic wind and engine noise up to volumes typical of general aviation airplanes. The control yoke (elevator -4 CDI course-deviation indicator and ailerons) is hydraulically loaded to provide the CRT cathode-ray tube appropriate variable force gradients. Rudder-pedal 1 force feel is supplied with springs. This simulator a DME distance measuring equipment has been used for related studies in automation and ECDI electronic course-deviation indicator displays (ref. 2). FREQ navigation receiver radio frequency A mathematical model for a typical high-wing, four-seat, single-engine, general aviation airplane was HSI horizontal-situation indicator used in the ATAS simulation. This model included IAF initial approach fix changes in flight-control effectiveness and force gra- IFR instrument flight rules dients as a function of airspeed, wing-flap-extension effects, an atmospheric wind-turbulence model, and a IJS instrument landing system radio navigation-aid data base. A landing gear model MANUAL ATAS operating mode that uses permitted touchdown and roll-out. manually input parameters to guide The simulation navigation-aid data base (ref. 4) the aircraft permits definition of a real navigation environment MDA minimum descent altitude for non- so that a subject pilot can fly cross-country flights and conduct instrument approaches by using stan- precision approach dard instrument charts. This data base includes the MSA minimum safe altitude location, Morse code audio identifier, and frequency, NAV VOR navigation receiver as applicable, of VOR, DME, NDB, marker beacon, localizer, and glide-slope transmitters. NDB nondirectional beacon The simulator is interfaced with a graphics com- NM nautical mile puter and a visual landing display system (VLDS). OBS omnibearing selector The graphics computer provides the capability to simulate advanced displays or alphanumeric data on PHF identifier for Patrick Henry Interna- cockpit cathode-ray tubes (CRT’s). The VLDS uses tional Airport a 1:750 scale terrain model and a closed-circuit tele- SBY standby mode vision to provide an out-the-windshield view during approach and landing. Realistic cloud-breakout and VLDS visual landing display system low-visibility effects are also provided. VOR very high frequency omnirange Symbols Experimental System h airplane barometric altitude The ATAS concept was designed to manage an h time rate of change of barometric altitude airplane autopilot, an autothrottle, and navigation radios for the pilot during an instrument approach V airplane airspeed from terminal area entry to the missed approach $ airplane heading angle point and, if necessary, through the missed approach. The ATAS performed this management task on the Test Equipment and Procedures basis of stored instrument approach data for each of the approaches to be flown. The purpose of this Airplane Simulation ATAS was to provide a means to study a concept for The current study was performed using the Lang- improving the pilot interface with high levels of cock- ley General Aviation Simulator. This research simu- pit automation. This interface improvement concept lator consists of an enclosed cockpit (fig. 1) interfaced was intended to reduce the number of avionics system 2 inputs (i.e., navigation radio tuning and autopilot heading channel is off, then the indicated heading mode changes) and the throttle control required of on the ATAS display tracks the airplane heading and the pilot during an approach, to improve situational movement of the course knob has no effect. If the status feedback to the pilot, and to reduce the pos-